Appendix C
Why Polymers Are More Susceptible Than Other Materials to Environmental Damage
S.S. Sternstein
Polymer matrix composites are highly susceptible to damage when used in various environmental extremes. This susceptibility results from a number of physical factors common to polymers in general and to the interface (or interphase, as appropriate) between the matrix and the reinforcing fibers.
Cohesive energy density, or intermolecular binding energy density, is a primary factor for polymer matrixes. This energy is very low in polymers compared to metals or ceramics. In polymers, the intermolecular forces binding polymer molecules together are classified as weak, meaning that these forces arise from bonds that are typically less than 5 kilocalories per mole. These include London dispersion forces,1 hydrogen bonds, and an assortment of dipolar and induced-dipolar intermolecular van der Waals forces that govern the interactions between two noncovalently bound atoms or molecules. These weak bonds are classified as secondary bonds when compared, for example, to the strong primary (20 to 80 kcal/mole) intermolecular bonds within metals or ceramics or to the intramolecular carbon-carbon backbone bonds holding the polymer molecule together. Note that these are also weak when compared to chemical crosslinking in a thermoset polymer matrix.
This structure, with very stiff (strong) bonding along the chain and very soft (weak) bonding among chains, leads to anisotropy that is observed in the polymeric matrix. The combination of weak and strong bonding leads to the polymer's complex deformation and flow behavior. Very long polymeric chains cannot simply translate past each other, as is the case for metallic atoms, but rather must allow each other to interdiffuse in a cooperative manner. This behavior is described by the term “reptation,” as it is symbolic of the way that snakes (reptiles) move across a surface. The ability to reptate is a very strong function of the ability of the chain to change its conformation, and this in turn is a strong function of temperature, because temperature governs the weak intermolecular forces.
For polymers, then, the weak intermolecular forces control chain conformation and therefore determine physical properties. This is especially important when interactions with the environment are considered, including moisture and other chemical species. In polymers, entropy changes associated with chain conformation are thermodynamically comparable with internal energy changes. This situation is seen in metals or ceramics only at very high temperatures—for example, at the melting point or at the glass transition temperature.
The weak intermolecular bonding in polymers leads to a correspondingly higher homologous temperature, which is the ratio of actual absolute temperature divided by the absolute glass transition temperature. In other words, a polymer at room temperature displays physical characteristics that are relatively speaking much closer to glass transition behavior than an inorganic glass at room temperature. In the case of inorganic silicate glasses, the strong intermolecular forces lead to glass transition temperatures that are typically many hundreds of degrees higher than typical glass transition
temperatures in the weakly bonded polymers. Indeed, attempts to synthesize polymers with higher glass transition temperatures (e.g., above 250°C) usually involve (1) very stiff backbone polymers with greatly reduced conformational mobility or (2) very highly crosslinked polymers, which effectively replace weak intermolecular bonds with strong (covalent) intermolecular bonds. In either case, polymers with higher glass transition temperatures tend to behave more like inorganic glasses than polymeric glasses, primarily because of the polymer molecule’s reduced ability to conform, leading to a reduced or eliminated ability to reptate.
Semicrystalline polymers follow the same pattern—that is, these polymers at room temperature are at much higher homologous temperatures than are metals at room temperature. Metals, of course, have much higher melting points owing to their higher intermolecular binding energies relative to polymers. Therefore, whether they are semicrystalline, amorphous, or crosslinked, polymers display behavior that is consistently representative of high homologous temperatures when compared to metals or ceramics.
For metals, high-temperature behavior (e.g., high-temperature creep) typically refers to behavior at temperatures higher than 0.5 of the melting point. It should be no surprise then that polymers at room temperature are already at high homologous temperature insofar as their behavior is concerned. For polymers, interactions that do not occur at low temperatures are to be expected at room temperature. Conversely, in the case of metals vacancies are known to exist at all temperatures but have little interactive effect at room temperature. Only at high temperatures is the vacancy concentration sufficient to allow edge dislocations to climb (move perpendicular to the slip plane). This climb is the result of the diffusion of atoms at the edge of the dislocation into adjacent vacancy sites. The result is higher dislocation mobility and greatly increased creep rates. At lower temperatures, the required diffusion process is too slow to provide significant climb, and this eliminates the enhanced creep rate. At room temperature, climb is essentially nonexistent and the interactions between vacancy concentration and dislocation mobility vanish. Thus a metal at room temperature exhibits “low” homologous temperature behavior.
A somewhat analogous situation exists in polymers when the glass transition temperature is approached, in that the greatly increased conformational diffusion rate leads to increased reptation rates. It should therefore be no surprise that the diffusion rate of small foreign molecules (e.g., from the environment) through the polymer is greatly increased as the glass transition temperature is approached or exceeded. In effect, the increased mobility of the polymeric chains provides the equivalent of “vacancies,” which facilitate the diffusion of the small foreign molecules. This is analogous to high-temperature behavior in a metal except that the polymer may actually be at room temperature! This further illustrates the concept that the sensitivity of polymer matrix composites to their environment is largely a result of the weak intermolecular forces and the resulting high homologous temperatures.2
To summarize, the weak intermolecular forces that are responsible for the unique physical properties of polymers (in the present context, most notably their low densities and viscoelastic behavior, including rate dependence, toughness, and ductility) are also responsible for their high sensitivity to their environment. This is illustrated by the fact that high homologous temperatures lead, in general, to conformational entropy related motions that are strongly dependent on temperature and external perturbations such as stress or the presence of foreign diffusing species. The interactive processes that can occur at high homologous temperatures complicate the mechanism-based, predictive modeling of extreme environmental effects on polymer-based composite performance. In effect, the noninteractive “low” homologous temperature regime is rarely encountered in polymer-based composites, and this clearly complicates the environmentally induced physical processes and their modeling. In turn, this makes the proper modeling of the interface even more important, since the complex behavior of the
matrix by itself may tend to obscure the role (or degradation) of the interface itself. Consequently, deconvolution of observed environmental phenomena into component matrix and interface phenomena may be severely complicated.
Finally, the dominant role played by conformational entropy in the structure and properties of high polymers leads to a huge multiplicity of morphological states having small differences in internal energy but huge differences in conformational entropy, as exemplified by heat-activated shrink wrap used in the electronics industry. This deformation recovery mechanism is entropy driven, with the polymeric chains going spontaneously upon heating from a low entropy state (prestretched and unshrunk) to a higher entropy state (shrunk). In the context of polymeric matrices for composites, the enormous range of morphological states leads to extreme sensitivity of the polymer matrix composite to its processing history and to a concomitant variability in its response to extreme environments. Thus, the “initial,” or baseline, properties of the composite may contain the “entropic memory” of the deformation history to which the matrix polymer was subjected during fabrication of the composite. Once again, this is comparable to the effects of processing history (cold forming, forging, hot drawing) of steel on subsequent properties. However, polymers display even more memory of prior history and a far greater range of morphologies, owing to the importance of conformational entropy.
An excellent example is given by polypropylene, which, depending on tacticity (atactic, syndiotactic, isotactic) and the distribution of tacticity, can exhibit morphologies ranging from completely amorphous to nearly completely crystalline, all at room temperature and without any change in chemical composition! Similar ranges of morphology can be obtained by the incorporation of small amounts of copolymer such as polyethylene. Any effort to develop a database or predictive model for polymer-based composites that does not consider the starting morphology of the polymeric matrix (including effects of prior deformation or processing history) is unlikely to be successful. Once again, recognition of the multidisciplinary character of this activity is scientifically compelling and essential for success.
As is the case with most engineering situations, advantages must be balanced with liabilities. The enormous advantages of polymer-based composites carry with them the added complexities and sensitivities to environmental extremes. Notwithstanding the difficulties, the potential payoff and advantages of environmentally stable, polymer-based composites makes long-term investigations of the type proposed herein not only appropriate but also essential to future engineering missions.
